Türkiye will soon launch the development of a real-time map of air pollution caused by vehicle traffic to draw realistic data and calculate emissions.
Yıldız Technical University, Environmental Engineering Department associate lecturer Levent Kuzu said: "Normally, air pollution estimations are made by generalization. In our method, there will be instantaneous, real data. We will try to make the most realistic estimation with the obtained data. Vehicles passing through the street will be classified one by one according to their types and traffic density."
Explaining the details of the TÜBITAK-supported project, Kuzu stated that the most important factor affecting air pollution in metropolitan cities are the fuels used in vehicles and for heating households, and those polluting gases and particles are released into the atmosphere depending on the content of the fuel.
Emphasizing that there are pollutants such as sulfur dioxide and nitrogen dioxide in gases that affect upper respiratory tract diseases, Kuzu underscored that there are critical values for these that should not be exceeded. "These vary daily or hourly depending on the type of pollutant. Our limit values are compatible with the European Union," he noted.
"Similar air pollutant problems are encountered in two different metropolitan cities – Istanbul and London. Particles smaller than 10 microns and nitrogen oxides are among the important pollutants. Combustion activities are effective in nitrogen oxides. High nitrogen oxide concentrations occur because of industry, heating and internal combustion engines in vehicles," he said, adding that vehicles and engines emit more emissions as their volumes increase.
"We refer to their emission figures as 'emission factors.' These factors vary depending on the age and size of the vehicle," he added.
Stating that there are air quality measurement stations at various points in Istanbul, he underlined that "the measurement we make differs from one another as placing measurement stations on every street is very difficult at the moment. In order to make a realistic estimation, we develop the air pollution estimation as close to the real value."
"We wanted to count the vehicles here using a deep learning method, counting from the camera and then sorting them into types, allowing us to obtain realistic data and calculate emissions," he said.
"Vehicles crossing the street will be classified one by one according to their types, and an emission calculation will be determined by calculating traffic density and waiting time in traffic. With the model, the concentrations of harmful gases and particles released into the atmosphere by the vehicles will be determined instantly. Once the project is completed, the model method developed can be used in the desired region. Normally, air pollution estimates are made by generalization. In our method, there will be instantaneous, real data," he shared.
Kuzu said that they aim to complete the project in two years and Beşiktaş Barbaros Boulevard will be a pilot area because of the heavy traffic, urbanization and presence of an air quality measurement station.
He further said that they would compare the data they recorded with the data of the measurement station in the region and ensure that the model reaches the optimum shape. "The important thing is to detect the pollutant accurately. After making the correct determination, information like fingerprints will be accessed and necessary measures will be taken," Kuzu concluded.
